Duke AI Health

What Duke AI Health does

Duke AI Health is a multidisciplinary, campus-spanning research initiative within Duke University School of Medicine that connects, strengthens, and amplifies theoretical and applied research in artificial intelligence and machine learning to address challenges in medicine, healthcare delivery, and population health. It operates a platform-plus-programs architecture that includes the Analytics Hub (investigator-to-methodologist matchmaking), Community of Practice (engagement and experiential learning), Data Science Fellowship Program, Proposal Studios, and Machine Learning Schools, alongside sub-brands such as the Collaborative to Advance Clinical Health Equity (CACHE) and the Spark Initiative for AI in Medical Imaging. The portfolio of research tools is broad and technically deep: PREVENT cardiovascular risk equations, EHR Foundation Models, IRIS retrieval-augmented classification, DANNCE 3D pose estimation, ConPLex/CoNCISE drug-target screening, Duke-PMA adolescent mental health risk model, the Snowball contact tracing platform, and the VISION-ACT mobile app, with several artifacts released as open-source code on GitHub.

The business model is that of an academic research initiative rather than a commercial software vendor. The organization is funded through Duke institutional budgets, philanthropy, and competitive federal and foundation grants (NIH, PCORI, CDC, AHA), with additional modest event revenue from Machine Learning Schools ($25-$275 per registrant) and the Innovation Summit ($150-$375). Go-to-market is community-led, distributing services through direct engagement with Duke investigators, regional clinical implementation projects with Duke Health, and a recurring international events program reaching Duke Kunshan and Duke NUS. Customer base is heterogeneous, encompassing Duke faculty and trainees, Duke Health clinical departments, federal agencies (CDC, NIH, PCORI), and external academic and industry partners (AHA, Brigham and Women's, UNC, Truveta, J&J, Globant, IBM, NVIDIA via MARK III).

Name

Duke AI Health

Legal name

Duke AI Health

Website

https://aihealth.duke.edu

Company type

Private

Operating status

Operating

Headcount range

11–50 employees

Short description

Duke AI Health is a multidisciplinary research initiative within Duke University School of Medicine that develops and applies AI and machine learning methods to medicine, serving Duke investigators, trainees, federal agencies, and external academic and industry partners through research, fellowships, and educational programs.

Ownership category

akta.pro rank

Product category

Academic Health AI Research

NAICS

Research and Development in Biotechnology (except Nanobiotechnology) (541714), Research and Development in the Physical, Engineering, and Life Sciences (except Nanotechnology and Biotechnology) (541715)

SIC

Services-Commercial Physical & Biological Research (8731), Services-Health Services (8000)

akta.pro primary industry

Health Informatics & Digital Health Programs (EDAHAGAJ)

akta.pro secondary industries

Biomedical Sciences & Medical Research Institutes (University-Based) (EDAHAGAI), Regulatory Readiness & Audit Automation (e.g., EU AI Act, NIST AI RMF, ISO/IEC 42001) (HDAAAMAE), Bias, Fairness & Non-Discrimination Testing (HDAAAMAC)

Keywords

• Health AI research
• Machine learning healthcare
• Clinical decision support
• Health data science
• Medical imaging AI

Headquarters

HQ city

Durham

HQ country

United States

HQ region

North America

Offices1 record

Markets served

GTM type

B2B

Offering type

Services

Cost components

Personnel, Technology or R&D, Operations, Infrastructure, Marketing or Sales

Revenue model

1. Institutional Funding: As a Duke University School of Medicine initiative, Duke AI Health is funded through institutional budgets, philanthropy, and government grants including NIH, PCORI, CDC, and AHA funding
2. Workshop and Event Registration: Registration fees for educational programs including Machine Learning Spring School ($275 general, $60 nonprofit, $25 student), Summit registration ($375 general, $150 nonprofit, free for students)
3. Grant-Funded Research: Research activities funded by federal agencies (NIH/NINDS, NIH/NEI, NIH/NICHD, PCORI, CDC) and foundations for specific project implementation

Pricing tiers

Go-to-market motion1 record

Distribution channels4 records

Marketing channels8 records

Core offering

Duke AI Health is a multidisciplinary academic research initiative at Duke University School of Medicine that performs foundational and applied AI/ML research for healthcare, provides analytics support and training to Duke investigators, and develops open-source research tools including EHR foundation models, clinical risk prediction algorithms, medical imaging AI, and behavioral analysis software. It operates structured programs such as the Analytics Hub, Community of Practice, Data Science Fellowship, Proposal Studios, Machine Learning Schools, and sub-initiatives including CACHE (health equity) and Spark (medical imaging).

Product overview

Duke AI Health is a comprehensive research and education initiative operating as a platform-plus-programs architecture within Duke University School of Medicine. The core organization connects AI/ML researchers across multiple schools and departments, offering structured programs including the Analytics Hub for researcher support, Community of Practice for engagement and learning, Data Science Fellowship for training, Proposal Studios for grant development, and Machine Learning Schools for education. Key sub-initiatives include CACHE (Collaborative to Advance Clinical Health Equity) for health equity work and Spark for medical imaging AI. The portfolio encompasses research tools including DANNCE for behavioral analysis, IRIS for clinical note interpretation, drug discovery pipelines (ConPLex, CoNCISE), risk prediction models (PREVENT, Duke-PMA), the Snowball contact tracing platform, and VISION-ACT mobile app. Duke Health AI Studio provides protected compute infrastructure for users. The organization also hosts major events including the Duke Summit on AI for Health Innovation, FAIR HEALTH Symposium, and educational workshops.

Differentiator

Problem solved

Functional benefit

Brands

• CACHE (Collaborative to Advance Clinical Health Equity): A program focused on advancing clinical health equity through AI and data science approaches
• Spark Initiative for AI in Medical Imaging
• Duke Machine Learning Schools

Products and services

• Machine Learning Schools Multi-day educational programs providing fundamentals of machine learning and AI methods with health applications, including hands-on practical sessions. Open to academic and industry participants, with international editions at Duke Kunshan University and Duke NUS Medical School. Registration: $275 standard, $60 nonprofit members, $25 students.
• Duke Summit on AI for Health Innovation Annual three-day summit fostering community of practice around health-oriented AI development, bridging clinical insight, engineering expertise, and industry perspectives. Includes panel discussions, fireside chats, design thinking sessions, and networking. Registration: $375 standard, $150 nonprofit, free for students.
• FAIR HEALTH Symposium Symposium and workshop series focused on algorithmic bias and clinical decision making in health AI, providing insight into sources of bias, ethical and legal considerations, and bias mitigation strategies.
• Analytics Hub Connects Duke researchers with quantitative methodologists for collaboration on health data science projects, working in partnership with CTSI BERD Methods Core. Researchers submit requests via the BERD Resource Request Form.
• Data Science Fellowship Program A 2-year training program in data science with direct application for healthcare, directed by Matt Engelhard, training the next generation of health data scientists with methodological rigor and healthcare relevance.
• Proposal Studios Structured program connecting investigators with data science experts to develop effective proposals for health data science research projects, including design workshops and proposal reviews. Launched in 2021 in collaboration with Duke +Data Science program.
• Community of Practice Engagement platform and experiential learning program for machine learning, AI, and health data science practitioners at Duke and beyond, directed by Shelley Rusincovitch and Ricardo Henao. Includes virtual seminars, scientific writing courses, and EHR study design workshops.
• Snowball Contact Tracing Platform Open-source cloud platform supporting respondent-driven sampling for COVID-19 contact tracing with EHR integration, participant-facing applications, consent/enrollment/peer recruitment workflows, and network analysis capabilities. Available on GitHub (duke-crucible/Snowball-API and duke-crucible/Snowball-UI).
• DANNCE (3-D Aligned Neural Network for Computational Ethology) Open-source ML tool for mapping precise 3D locations of body joints on animals from video for behavioral research. Developed by Timothy Dunn's team.
• Duke Health AI Studio AI platform environment offering various AI platforms for Duke users in a protected compute environment, supported by Duke Health Technology Solutions (DHTS).
• CACHE (Collaborative to Advance Clinical Health Equity) Collaborative initiative focused on developing and implementing AI and health data science solutions to advance clinical health equity across Duke health systems and beyond. Includes multiple RFAs and clinical projects on maternal morbidity, prostate cancer screening, and colorectal cancer screening.
• Spark Initiative for AI in Medical Imaging Duke-wide initiative where technical and clinical experts collaborate to develop, evaluate, and implement algorithms for medical imaging, including X-rays, CT scans, and other radiological images. Directed by Maciej Mazurowski.

Named customers6 records

Segments4 records

Ideal customer profiles4 records

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

AI capability13 records

Feature6 records

Partnerships

14 partnerships are on record, tiered core and supporting.

• American Heart AssociationcoreStrategic or Co-development PartnerMulti-year collaborative research with AHA researchers on stroke risk prediction modeling, harmonizing data from Framingham, MESA, ARIC, and REGARDS cohort studies, and developing machine learning-based risk prediction models. NIH/NINDS R61-NS120246 funded project.
• Duke Faculty Advancement Seed Grants ProgramsupportingStrategic or Co-development PartnerSPARC Initiative for Algorithmic Bias Detection and Mitigation supported by Duke Faculty Advancement Seed Awards, bringing together Duke School of Nursing faculty and AI Health colleagues to develop training for eliminating bias in clinical algorithms.
• Duke Autism Center of Excellence (ACE)coreStrategic or Co-development PartnerCollaboration on autism screening research using EHR data and developing clinical decision support prototypes for early identification of autism and co-occurring conditions. Funded by NICHD P50HD093074.
• Brigham and Women's HospitalcoreStrategic or Co-development PartnerCollaborative research on cardiovascular disease risk prediction for women with adverse pregnancy outcomes, part of NIH/NHLBI 1R01HL167960-01A1 grant.
• University of North Carolina-Chapel HillcoreStrategic or Co-development PartnerMulti-site collaborative research on cardiovascular disease risk prediction combining data from Duke, UNC, and Nurses' Health Study.
• TruvetasupportingTechnology or IntegrationTruveta provides electronic health record data for research, including PREVENT cardiovascular risk equation evaluation across 550,000+ adults in multi-system dataset.
• Johnson & JohnsonsupportingGTM or Marketing PartnerSummit sponsor for the Duke Summit on AI for Health Innovation, along with other industry partners.
• Globant - Healthcare & Life SciencessupportingGTM or Marketing PartnerSummit sponsor providing support for the Duke Summit on AI for Health Innovation.
• IBM Data AI and AutomationsupportingGTM or Marketing PartnerSummit sponsor for the Duke Summit on AI for Health Innovation.
• MARK III Systems with NVIDIAsupportingGTM or Marketing PartnerSummit sponsor for the Duke Summit on AI for Health Innovation.
• Coalition for Health AI (CHAI)coreStrategic or Co-development PartnerDuke is a founding partner of CHAI, whose mission is to increase trustworthiness of AI by developing guidelines to drive high-quality health care through adoption of credible, fair, and transparent health AI systems. Duke AI Health Director Michael Pencina co-founded and co-chairs CHAI.
• Duke Health AI Evaluation and Governance (E&G) ProgramcoreStrategic or Co-development PartnerDevelops methods and frameworks for testing and overseeing health AI applications. E&G program serves as incubator from 2021 until launch in 2024, now at healthaigovernance.duke.edu.
• CTSI Biostatistics, Epidemiology, and Research Design (BERD) Methods CorecoreStrategic or Co-development PartnerAI Health collaborates closely with BERD to help connect researchers with appropriate quantitative methodologists for collaboration. Submit requests via BERD Resource Request Form.
• Duke +Data Science (+DS) ProgramcoreStrategic or Co-development PartnerPartnership with +DS program for Proposal Studios and educational outreach. +DS has supported 115 co-curricular learning experiences, 8 machine learning schools, and 72 students in applied projects since 2018. Their Coursera 'Introduction to Machine Learning' has reached 129,000+ learners.

Scale indicators7 records

Recent moves8 records

Expansion highlights6 records

Others

• Coalition for Health AI (CHAI): CHAI is the national multi-stakeholder consortium co-founded and co-chaired by Duke AI Health's director, focused on guidelines and guardrails for trustworthy health AI. It is not a direct research competitor but an ecosystem consortium where Duke AI Health exerts outsized influence alongside other academic and industry members.

Direct peers

• Mount Sinai Hasso Plattner Institute for Digital Health: Mount Sinai's Hasso Plattner Institute for Digital Health is an academic medical center-based AI research and engineering organization working on EHR-based machine learning, clinical NLP, and deployed digital health tools. Like Duke AI Health, it combines research, training, and clinical implementation within a single integrated structure.
• Stanford Institute for Human-Centered Artificial Intelligence (HAI): Stanford HAI is a multidisciplinary AI research institute with a dedicated Health & Medicine focus area covering EHR foundation models, clinical decision support, and AI policy. Like Duke AI Health, it operates as a campus-spanning initiative that convenes medicine, engineering, and policy faculty, hosts educational programs, and engages in national AI governance discussions.
• Harvard Medical School Department of Biomedical Informatics: Harvard DBMI conducts AI/ML research on EHR data, clinical NLP, and translational bioinformatics and runs training programs and collaborations with Harvard-affiliated hospitals. It is one of the most established academic biomedical informatics groups and a direct comparator to Duke AI Health in scope and mission.
• Penn Medicine Center for AI in Healthcare (CAIH): Penn CAI conducts applied AI research across imaging, EHR, and clinical workflows within a top-tier academic medical center, with explicit training, community engagement, and implementation programs that closely parallel Duke AI Health's Analytics Hub, Spark, and Fellowship offerings.
• UCSF Bakar Computational Health Sciences Institute: UCSF's Bakar Institute drives translational computational health research, including EHR-based predictive modeling, NLP for clinical text, and AI for clinical decision support. It mirrors Duke AI Health's combination of faculty-led research programs, training fellowships, and industry partnerships anchored at an academic medical center.
• Johns Hopkins Malone Center for Engineering in Healthcare: Johns Hopkins' Malone Center is a multidisciplinary initiative that brings together engineering, medicine, and computer science to develop AI and data science methods for clinical care. Like Duke AI Health, it emphasizes collaborations across schools, training programs, and translation of research into clinical practice.

Broad incumbents

• Mayo Clinic Platform: Mayo Clinic Platform is a large-scale, multi-product health AI initiative inside Mayo Clinic offering data, analytics, and validated AI model deployment to external partners. It overlaps with Duke AI Health's predictive modeling and clinical decision support research but operates as a broader commercial-facing portfolio with industry partnerships.

Market position

Strengths4 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/duke-ai-health

Revenue estimate

Valuation estimate

Number of profiles

Profiles9 records

Funding overview

Funding rounds

Investors

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Duke AI Health